Validity of anthropometric equations to estimate infant fat mass at birth and in early infancy

BACKGROUND In newborns and children, body fat estimation equations are often used at different ages than the age used to develop the equations. Limited validation studies exist for newborn body fat estimation equations at birth or later in infancy. The study purpose was to validate 4 newborn fat mass (FM) estimation equations in comparison to FM measured by air displacement plethysmography (ADP; the Pea Pod) at birth and 3 months. METHODS Ninety-five newborns (1-3 days) had their body composition measured by ADP and anthropometrics assessed by skinfolds. Sixty-three infants had repeat measures taken (3 months). FM measured by ADP was compared to FM from the skinfold estimation equations (Deierlein, Catalano, Lingwood, and Aris). Paired t-tests assessed mean differences, linear regression assessed accuracy, precision was assessed by R2 and standard error of the estimate (SEE), and bias was assessed by Bland-Altman plots. RESULTS At birth, FM measured by ADP differed from FM estimated by Deierlein, Lingwood and Aris equations, but did not differ from the Catalano equation. At 3 months, FM measured by ADP was different from all equations. At both time points, poor precision and accuracy was detected. Bias was detected in most all equations. CONCLUSIONS Poor agreement, precision, and accuracy were found between prediction equations and the criterion at birth and 3 months.